Skip to content

Targets and horizons

RokGrah edited this page Feb 14, 2023 · 11 revisions

There is no obligation to submit for all suggested targets or horizons, and it is up to you to decide what you are comfortable doing with your model.

Please take care when submitting to be clear which target and horizon you are using: use this page and the formatting guide.

Targets

You can submit projections for only one, or multiple targets. We only use incident counts per week: the count of new cases or deaths per week. We do not use cumulative (running total) forecasts (this varies from other hubs).

Key targets:

  • Incident infections
  • Incident deaths
  • Incident hospital admissions
  • Administered vaccine doses [if relevant for a given scenario round]

Additional targets that might be added:

  • Incident cases: only if incident infections cannot be modelled
  • Intensive care (ICU) admissions [target removed as of Round 3]

Horizon and frequency

Horizon

For the scenario hub, we require projections to include at least 16 weeks and up to 520 weeks ahead, depending on the scenario round.

Date format

Horizons should use the Epidemiological Week (EW) format, defined by the US CDC. Each week starts on Sunday and ends on Saturday. For example:

  • A 1 week ahead projection of incident deaths submitted on Monday 4th January 2021 should cover the seven days including Sunday 3rd January to Saturday 9th January. In this example, the relevant fields would have these values:
    • target_variable = "inc death"
    • horizon = "1 wk"
  • A 4 week ahead case projection submitted on Monday 4th January would include the period Sunday 24th January to Saturday 30th January 2021. In this example:
    • target_variable = "inc death"
    • horizon = "4 wk"

Location

You can submit for a single country, or any combination of the target countries (EU, EFTA and the UK). We use ISO-2 codes to identify countries.

Teams can also (or only) submit results for a "Hypothetical" location, for example where parameters are set as the average of multiple countries.

Currently we do not accept sub-national projections. Sub-national country datasets can be incompatible with each over and over time, making comparisons between projections unclear.